Update app/streamlit_app.py

app/streamlit_app.py

@@ -977,7 +977,7 @@ def render_prediction_interface():
                     st.error(validation_message)
                 else:
                     # Make prediction with enhanced UI
-                    result = app.make_prediction_request(user_text)
+                    result = app_manager.make_prediction_request(user_text)
 
                     if 'error' in result:
                         st.error(f"Analysis failed: {result['error']}")
@@ -1032,7 +1032,7 @@ def render_prediction_results(result: Dict, user_text: str, show_advanced: bool)
 
     with col1:
         # Advanced confidence gauge
-        fig_gauge = app.create_advanced_confidence_gauge(confidence, prediction)
+        fig_gauge = app_manager.create_advanced_confidence_gauge(confidence, prediction)
         st.plotly_chart(fig_gauge, use_container_width=True)
 
     with col2:
@@ -1119,8 +1119,8 @@ def validate_text_input(text: str) -> tuple[bool, str]:
     if len(text) < 10:
         return False, "Text must be at least 10 characters long for reliable analysis."
 
-    if len(text) > app.config['max_text_length']:
-        return False, f"Text must be less than {app.config['max_text_length']:,} characters."
+    if len(text) > app_manager.config['max_text_length']:
+        return False, f"Text must be less than {app_manager.config['max_text_length']:,} characters."
 
     # Check for suspicious content
     suspicious_patterns = ['<script', 'javascript:', 'data:', 'onclick=']
@@ -1160,7 +1160,7 @@ def render_batch_analysis():
                 with col1:
                     max_articles = st.number_input("Max articles to process", 
                                                  min_value=1, 
-                                                 max_value=min(len(df), app.config['max_batch_size']),
+                                                 max_value=min(len(df), app_manager.config['max_batch_size']),
                                                  value=min(len(df), 50))
                 
                 with col2:
@@ -1184,7 +1184,7 @@ def render_batch_analysis():
                         status_text.text(f"Processing article {i+1:,}/{len(df):,}...")
                         progress_bar.progress((i + 1) / len(df))
 
-                        result = app.make_prediction_request(row['text'])
+                        result = app_manager.make_prediction_request(row['text'])
 
                         if 'error' not in result:
                             results.append({
@@ -1325,7 +1325,7 @@ def render_analytics_dashboard():
     if st.session_state.prediction_history:
         st.subheader("Prediction History Analysis")
         
-        fig_history = app.create_prediction_history_chart()
+        fig_history = app_manager.create_prediction_history_chart()
         if fig_history:
             st.plotly_chart(fig_history, use_container_width=True)
 
@@ -1410,7 +1410,7 @@ def render_cv_results_dashboard():
     st.markdown("*Statistical validation and model performance analysis*")
 
     # Get CV results
-    cv_results = app.get_cv_results_from_api()
+    cv_results = app_manager.get_cv_results_from_api()
     
     if cv_results and 'error' not in cv_results:
         # Model information header
@@ -1431,7 +1431,7 @@ def render_cv_results_dashboard():
         cv_data = cv_results.get('cross_validation', {})
         if cv_data:
             # Create comprehensive CV visualization
-            fig_cv = app.create_cv_performance_visualization(cv_results)
+            fig_cv = app_manager.create_cv_performance_visualization(cv_results)
             if fig_cv:
                 st.plotly_chart(fig_cv, use_container_width=True)
 
@@ -1799,7 +1799,7 @@ def render_health_monitoring():
         st.session_state.auto_refresh = auto_refresh
 
     # System health overview
-    app.render_system_health_dashboard()
+    app_manager.render_system_health_dashboard()
 
     # Environment information
     st.markdown("---")
@@ -1839,9 +1839,9 @@ def render_health_monitoring():
             st.markdown(f"{status} {name}")
 
     # Model performance monitoring
-    if app.api_available:
+    if app_manager.api_available:
         try:
-            metrics_response = app.session.get(f"{app.config['api_url']}/metrics", timeout=10)
+            metrics_response = app_manager.session.get(f"{app_manager.config['api_url']}/metrics", timeout=10)
             if metrics_response.status_code == 200:
                 metrics_data = metrics_response.json()
                 
@@ -1920,7 +1920,7 @@ def render_realtime_metrics():
         metrics_timeframe = st.selectbox("Timeframe", ["Last Hour", "Last 4 Hours", "Last 24 Hours"])
 
     # Real-time monitoring data
-    monitoring_data = app.get_monitoring_metrics_from_api()
+    monitoring_data = app_manager.get_monitoring_metrics_from_api()
     
     if monitoring_data:
         # Current performance metrics